Among the 2007 strategic goals of the National Cancer Institute (NCI) are improving early detection of cancers and overcoming cancer health disparities. To help NCI achieve the vision of "A Nation free from the suffering and death due to cancer by 2015 with dramatic reductions in cancer incidence," RTI International and the University of Illinois at Urbana-Champaign have partnered to advance the understanding of discrepancies in the participation in cancer screening across subpopulations and regions of the United States. The major aims are to (1) contribute to a better understanding of the factors that influence cancer screening rates and characteristics of locales with the lowest rates, so that these disparities might be reduced, and (2) advance new geospatial methods and measurements in social science and behavioral health research and apply them to the study of cancer screening rates. We will achieve this by: developing and making publicly available a database of geospatial measures, which include characteristics of people in places and characteristics of the places themselves; developing and making publicly available new spatial analysis software capabilities and a Web application, built to standards developed by the Open Geospatial Consortium, that allows analysis of data in the databases along with proprietary data linked to the system by the user;and using these data and tools to estimate multivariate models explaining variability in cancer screening across people and places, and disseminating the findings and other support materials, providing information that will guide other researchers who use these data and tools. In our research, we will examine determinants of screening for two types of cancer (breast and colorectal) among large retrospective cohorts of the elderly population residing in all areas of the United States. We will combine the information contained in detailed health administrative records (Medicare claims) with our newly constructed geospatial variables pertaining to contextual factors that vary across small areas of geography. Our research papers will contribute to a better understanding of the factors that cause variability in cancer screening, and show the significance of place-specific heterogeneity in assessing health behaviors and outcomes. We will demonstrate how this approach might be used to identify locales where targeted interventions could be most effective. Our research, readily available software, public-use geospatial database and Web portal with interactive mapping and geovisualization capabilities will advance the application of new geospatial methods and measurement in social science and behavioral health research.